Foreign Substance Detection via Image Smoothing and Difference Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The computational complexity of spatial frequency analysis for detecting regions with predetermined spatial frequency spectra in images poses a significant load on detection apparatuses, particularly in applications like vehicles and imaging systems, making it challenging to efficiently detect foreign substances like dirt or water droplets on camera lenses.

Innovation Solution

A detection apparatus and imaging system that uses a controller to generate smoothed images and difference images, allowing for the detection of low-frequency regions without analyzing spatial frequency spectra, thereby reducing computational load and enabling the detection of foreign substances by comparing captured images with their smoothed versions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If spatial frequency analysis is used to detect foreign substances on the lens surface, then detection accuracy is improved, but computational complexity increases significantly

Engineering Contradiction:
Improvedetection accuracyVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent divides the captured image into multiple blocks and performs frequency analysis on each block separately. This segmentation allows the system to detect foreign substances with high accuracy in specific regions without requiring computationally intensive full-image analysis, thus resolving the contradiction between detection accuracy and computational complexity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies frequency analysis only to specific blocks where foreign substances are suspected, rather than analyzing the entire image. By performing partial analysis on selected regions, the system achieves sufficient detection accuracy while significantly reducing the overall computational burden.

Inventive Principle:
Principle #16Partial or excessive action

2Reliability

If spatial frequency analysis is performed on the entire image, then foreign substance detection capability is improved, but processing time increases

Engineering Contradiction:
Improveforeign substance detection capabilityVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The image is divided into multiple blocks for parallel processing. This segmentation enables the system to maintain comprehensive foreign substance detection capability across the entire image while reducing processing time by distributing the computational workload across multiple smaller regions that can be analyzed simultaneously or selectively.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system performs frequency analysis on selected blocks rather than the entire image, achieving sufficient foreign substance detection capability with reduced processing time by focusing computational resources only on regions where foreign substances are likely to be present.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3493536B1Detection device, image pickup device, movable body, and detection method
Publication Date: 2022.10.12 KYOCERA CORP
  • EP3493536B1 patent drawingFigure 1
  • EP3493536B1 patent drawingFigure 2
  • EP3493536B1 patent drawingFigure 3

AI summary

A detection apparatus includes an image acquisition interface that acquires a captured image captured by an imaging unit and a controller that generates or acquires a smoothed image yielded by smoothing the captured image. The controller compares the captured image and the smoothed image and detects a low-frequency region having a predetermined spatial frequency spectrum from the captured image.